Quits per layoff
This chart divides the number of U.S. workers who quit their jobs by the number laid off or discharged, using BLS JOLTS data via FRED, smoothed over three months. A dashed line holds layoffs at their 2019 rate, so the gap between the lines shows how much layoffs, rather than quitting, move the ratio. It does not show how much total turnover there is.
What does it show?
Quits outnumber layoffs 1.8 to 1, a bit below 2019's 1.9; the shortfall comes from less quitting, while the layoffs rate is lower than in 2019.
Methodology
Quits divided by layoffs and discharges, U.S. total nonfarm, seasonally adjusted, from the BLS Job Openings and Labor Turnover Survey (JOLTS), retrieved through FRED (JTSQUL and JTSLDL, levels in thousands). The solid line is a trailing three-month ratio of sums, dated at the last month of its window. A value of 2 means two workers quit for every one laid off or discharged; below 1, layoffs outnumber quits. The ratio uses levels rather than the published rates, which are rounded to one decimal and would move the ratio by up to 0.19 in a single month. The dashed line shows what the ratio would have been if only quits had moved: the three-month quits rate divided by the 2019 average layoffs rate (1.21%). It is the quits rate rescaled, so the distance between the two lines is the part of the ratio that layoffs move. Both rates are rebuilt unrounded from levels over BLS CES total nonfarm employment (PAYEMS, retrieved through FRED), the JOLTS rate denominator; rounded, the rebuilt layoffs rate matches the published BLS rate in every month from 2016 to the latest. Other separations (retirements, deaths, transfers) are excluded. Quits include job-to-job moves, and layoffs and discharges include dismissals for cause. The ratio does not show how much turnover there is: the same value can come from many quits and many layoffs or from few of each. JOLTS begins in December 2000, so the history covers three recessions. JOLTS is a sample survey, so many month-to-month moves in either leg are within sampling error; the three-month window damps but does not remove this. The latest month is preliminary and is revised the following month, and annual revisions can also change earlier years. Data were retrieved from FRED on 2026-10-22. BLS.gov cannot vouch for data or analyses after retrieval from BLS.gov.